2.2.2 Sequential Quadratic Programming



next up previous contents index
Next: 2.3 Empirical Model Building Up: 2.2 Nonlinear Programming Previous: 2.2.1 Sequential Linear Programming

2.2.2 Sequential Quadratic Programming

  The sequential quadratic programming (SQP) code of [95] was also adapted and used as an alternative solution method. This code employs the Han Powell [82] variable metric method with the watchdog technique [18]. It replaces the objective with its quadratic approximation and the constraints with linear functions. The resultant quadratic programming problem is then solved to determine a feasible direction of descent that satisfies the active constraints. Once such a direction is found, a robust line search algorithm selects the size of the parameters update step.


Martin Stiftinger
Tue Aug 1 19:07:20 MET DST 1995